Search (1352 results, page 7 of 68)

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 3796) [ClassicSimilarity], result of:
      0.18536724 = score(doc=3796,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 3796, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=3796)
  0.25 = coord(1/4)

Abstract

State of the art review of the use of information sources by family doctors is important for both practical and theoretical reasons. Analysis of the ways in which family doctors handle information may point to opportunities for improvement. Such efforts lead to improvements in the methodology of literature research in general. Reports on a review of the literature on information use by family doctors. 11 relevant research publications were found. The data showed that family doctors used colleagues most often as information sources followed by periodicals and books. This outcome corresponded with results found for other professions. Several factors influenced the use of information sources by family doctors, including the physical, functional and intellectual accessibility of the sources; the doctor's age; participation by the doctor in research or education; the social context of the doctor; practice characteristics; and the stage of the informatio gathering process. The publications studied suggested ways to improve information gathering in the areas of computerization, education, library organization, and periodical articles

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 5863) [ClassicSimilarity], result of:
      0.18536724 = score(doc=5863,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 5863, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=5863)
  0.25 = coord(1/4)

Abstract

In data mining, we emphasize the need for learning from huge, incomplete, and imperfect data sets. To handle noise in the problem domain, existing learning systems avoid overfitting the imperfect training examples by excluding insignificant patterns. The problem is that these systems use a limiting attribute-value language for representing the training examples and the induced knowledge. Moreover, some important patterns are ignored because they are statistically insignificant. In this article, we present a framework that combines genetic programming and inductive logic programming to induce knowledge represented in various knowledge representation formalisms from noisy databases (LOGENPRO). Moreover, the system is applied to one real-life medical database. The knowledge discovered provides insights to and allows better understanding of the medical domains

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 1690) [ClassicSimilarity], result of:
      0.18536724 = score(doc=1690,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 1690, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=1690)
  0.25 = coord(1/4)

Abstract

The value of an organization is given not just by its tangible assets but also by the knowledge embodied in its employees and in its internal and external structures. Explicit knowledge can be shared as information and is easy to handle. Tacit knowledge has been neglected by effectiveness-orientied management techniques but is crucial for both the operational efficiency and the core competence of an organization. This book provided a survey of the use of information technology for knowledge management, and its chapters present specific research on how technologies such as computer-supported cooperative work (CSCW), workflow, and groupware can support the creation and sharing of knowledge in organizations

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 2850) [ClassicSimilarity], result of:
      0.18536724 = score(doc=2850,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 2850, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=2850)
  0.25 = coord(1/4)

Abstract

Answering open-domain factual questions requires Natural Language processing for refining document selection and answer identification. With our system QALC, we have participated in the Question Answering track of the TREC8, TREC9 and TREC10 evaluations. QALC performs an analysis of documents relying an multiword term searches and their linguistic variation both to minimize the number of documents selected and to provide additional clues when comparing question and sentence representations. This comparison process also makes use of the results of a syntactic parsing of the questions and Named Entity recognition functionalities. Answer extraction relies an the application of syntactic patterns chosen according to the kind of information that is sought, and categorized depending an the syntactic form of the question. These patterns allow QALC to handle nicely linguistic variations at the answer level.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 3561) [ClassicSimilarity], result of:
      0.18536724 = score(doc=3561,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 3561, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=3561)
  0.25 = coord(1/4)

Abstract

During the fall 2001 semester 15 first-year undergraduates were interviewed about their information-seeking behavior. Undergraduates completed a short-answer questionnaire, the Measure of Epistemological Reflection, measuring their epistemological beliefs and searched the Web and an online public access catalog using tasks from the Reflective Judgment Interview that assessed their reflective judgment level. Undergraduates talked aloud while searching digital environments about the decisions they were making about the information they encountered while transaction analyses software (Lotus ScreenCam) recorded both their search moves and their decision-making through verbal protocol analysis. Analyses included examining the relationship between undergraduates' epistemological beliefs and reflective judgment and how they searched for information in these digital environments. Results indicated that there was a relationship between epistemological beliefs and reflective judgment and information-seeking behavior. Undergraduates' at higher stages of epistemological development exhibited the ability to handle conflicting information sources and to recognize authoritative information sources.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 5330) [ClassicSimilarity], result of:
      0.18536724 = score(doc=5330,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 5330, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=5330)
  0.25 = coord(1/4)

Abstract

Purpose - During multitasking, humans handle multiple tasks through task switching or engage in multitasking information behaviors. For example, a user switches between seeking new kitchen information and medical information. Recent studies provide insights these complex multitasking human information behaviors (HIB). However, limited studies have examined the interplay between information and non-information tasks. Design/methodology/approach - The goal of the paper was to examine the interplay of information and non-information task behaviors. Findings - This paper explores and speculates on a new direction in HIB research. The nature of HIB as a multitasking activity including the interplay of information and non-information behavior tasks, and the relation between multitasking information behavior to cognitive style and individual differences, is discussed. A model of multitasking between information and non-information behavior tasks is proposed. Practical implications/limitations - Multitasking information behavior models should include the interplay of information and non-information tasks, and individual differences and cognitive styles. Originality/value - The paper is the first information science theoretical examination of the interplay between information and non-information tasks.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 513) [ClassicSimilarity], result of:
      0.18536724 = score(doc=513,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 513, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=513)
  0.25 = coord(1/4)

Abstract

When the Bahá'í World Centre Library was organized on professional principles in 1977, there was no system adequate to handle the classification of a large specialized collection on the Bahá'í Faith. The librarian, in consultation with knowledgeable Bahá'í is and Bahá'í librarians, began a staged development of a classification scheme that could be adapted to the uses of institutions with significant collections of Bahá'í material. Following the development of two preliminary expansions of Dewey and Library of Congress systems, and study of the types and subject matter of materials being produced, a reasonably complete classification for the current subject areas covered in Bahá'í publications was developed. This classification should be of benefit to academic libraries planning large acquisitions of Bahá'í materials, and particularly to the growing numbers of developing Bahá'í libraries.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 625) [ClassicSimilarity], result of:
      0.18536724 = score(doc=625,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 625, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=625)
  0.25 = coord(1/4)

Abstract

The cataloging of music sound recordings is both challenging and interesting. As the technologies used to produce sound recordings change, these changes must be reflected in both cataloging theory and practice. Three formats: analog disc, cassette tape, and compact disk, all of which are readily available on the market, present special challenges to catalogers who must consider what might be the most effective way of handling these materials following the AACR2 cataloging rules and interpretations from Library of Congress. This paper examines the actual cataloging of those formats as done by several institutions and raises questions such as how to handle these materials in ways which will eliminate redundancy and increase efficiency in the practice of cataloging. Finally, an alternative approach, drawing on AACR2 practice in other areas, is suggested.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 1718) [ClassicSimilarity], result of:
      0.18536724 = score(doc=1718,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 1718, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=1718)
  0.25 = coord(1/4)

Abstract

Mark Stapleton and Matt Adamson began their presentation by describing how Dow Jones' Factiva range of information services processed an average of 170,000 documents every day, drawn from over 10,000 sources in 22 languages. These documents are categorized within five facets: Company, Subject, Industry, Region and Language. The digital feeds received from information providers undergo a series of processing stages, initially to prepare them for automatic categorization and then to format them ready for distribution. The categorization stage is able to handle 98% of documents automatically, the remaining 2% requiring some form of human intervention. Depending on the source, categorization can involve any combination of 'Autocoding', 'Dictionary-based Categorizing', 'Rules-based Coding' or 'Manual Coding'

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 1991) [ClassicSimilarity], result of:
      0.18536724 = score(doc=1991,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 1991, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=1991)
  0.25 = coord(1/4)

Abstract

Multitasking is the human ability to handle the demands of multiple tasks. Multitasking behavior involves the ordering of multiple tasks and switching between tasks. People often multitask when using information retrieval (IR) technologies as they seek information on more than one information problem over single or multiple search episodes. However, limited studies have examined how people order their information problems, especially during their Web search engine interaction. The aim of our exploratory study was to investigate assigned information problem ordering by forty (40) study participants engaged in Web search. Findings suggest that assigned information problem ordering was influenced by the following factors, including personal interest, problem knowledge, perceived level of information available on the Web, ease of finding information, level of importance and seeking information on information problems in order from general to specific. Personal interest and problem knowledge were the major factors during assigned information problem ordering. Implications of the findings and further research are discussed. The relationship between information problem ordering and gratification theory is an important area for further exploration.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 3024) [ClassicSimilarity], result of:
      0.18536724 = score(doc=3024,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 3024, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=3024)
  0.25 = coord(1/4)

Abstract

Over time, researchers have acknowledged the importance of understanding the users' strategies in the design of search systems. However, when involving users in the comparison of search systems, methodological challenges still exist as researchers are pondering on how to handle the variability that human participants bring to the comparisons. This paper present methods for controlling the complexity of user-centered evaluations of search user interfaces through within-subjects designs, balanced task sets, time limitations, pre-formulated queries, cached result pages, and through limiting the users' access to result documents. Additionally, we will present our experiences in using three measures - search speed, qualified search speed, and immediate accuracy - to facilitate the comparison of different search systems over studies.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 3445) [ClassicSimilarity], result of:
      0.18536724 = score(doc=3445,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 3445, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=3445)
  0.25 = coord(1/4)

Abstract

This article explores how to develop complex data driven user models that go beyond the bag of words model and topical relevance. We propose to learn from rich user specific information and to satisfy complex user criteria under the graphical modelling framework. We carried out a user study with a web based personal news filtering system, and collected extensive user information, including explicit user feedback, implicit user feedback and some contextual information. Experimental results on the data set collected demonstrate that the graphical modelling approach helps us to better understand the complex domain. The results also show that the complex data driven user modelling approach can improve the adaptive information filtering performance. We also discuss some practical issues while learning complex user models, including how to handle data noise and the missing data problem.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 3453) [ClassicSimilarity], result of:
      0.18536724 = score(doc=3453,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 3453, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=3453)
  0.25 = coord(1/4)

Abstract

Author name disambiguation deals with clustering the same-name authors into different individuals. To attack the problem, many studies have employed a variety of disambiguation features such as coauthors, titles of papers/publications, topics of articles, emails/affiliations, etc. Among these, co-authorship is the most easily accessible and influential, since inter-person acquaintances represented by co-authorship could discriminate the identities of authors more clearly than other features. This study attempts to explore the net effects of co-authorship on author clustering in bibliographic data. First, to handle the shortage of explicit coauthors listed in known citations, a web-assisted technique of acquiring implicit coauthors of the target author to be disambiguated is proposed. Then, a coauthor disambiguation hypothesis that the identity of an author can be determined by his/her coauthors is examined and confirmed through a variety of author disambiguation experiments.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 3741) [ClassicSimilarity], result of:
      0.18536724 = score(doc=3741,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 3741, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=3741)
  0.25 = coord(1/4)

Abstract

Authorship attribution supported by statistical or computational methods has a long history starting from the 19th century and is marked by the seminal study of Mosteller and Wallace (1964) on the authorship of the disputed Federalist Papers. During the last decade, this scientific field has been developed substantially, taking advantage of research advances in areas such as machine learning, information retrieval, and natural language processing. The plethora of available electronic texts (e.g., e-mail messages, online forum messages, blogs, source code, etc.) indicates a wide variety of applications of this technology, provided it is able to handle short and noisy text from multiple candidate authors. In this article, a survey of recent advances of the automated approaches to attributing authorship is presented, examining their characteristics for both text representation and text classification. The focus of this survey is on computational requirements and settings rather than on linguistic or literary issues. We also discuss evaluation methodologies and criteria for authorship attribution studies and list open questions that will attract future work in this area.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 524) [ClassicSimilarity], result of:
      0.18536724 = score(doc=524,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 524, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=524)
  0.25 = coord(1/4)

Abstract

In the processes of information organization, information retrieval and information visualization one point in common can be found: they are strongly connected by the procedures associated to the indexing of texts or documents. Indexing is an essential component of text analysis, and the indexing process have equal importance both for retrieval and for the visualization of information. In this context, it is worth mentioning the solutions that use automatic indexing. The researches that proposes solutions for automatic indexing are based on different theoretical assumptions such as: statistics, linguistics and controlled vocabulary (Leiva 1999). Most solutions developed hybrid models combining these theoretical assumptions. Other solutions to the problem of automatic indexing are based on theories that allow the treatment of uncertainty, imprecision and vagueness. The aim of this paper is to argue the theoretical potential for use in hybrid models of automatic indexing, the paraconsistent logic, a non-classical logic, with capacity to handle situations that involve uncertainty, imprecision and vagueness.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 1860) [ClassicSimilarity], result of:
      0.18536724 = score(doc=1860,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 1860, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=1860)
  0.25 = coord(1/4)

Abstract

Since its emergence as a discipline, in the nineteenth century (1889), the theory and practice of Archival Science have focused on the arrangement and description of archival materials as complementary and inseparable nuclear processes that aim to classify, to order, to describe and to give access to records. These processes have their specific goals sharing one in common: the representation of archival knowledge. In the late 1980 a paradigm shift was announced in Archival Science, especially after the appearance of the new forms of document production and information technologies. The discipline was then invited to rethink its theoretical and methodological bases founded in the nineteenth century so it could handle the contemporary archival knowledge production, organization and representation. In this sense, the present paper aims to discuss, under a theoretical perspective, the archival representation, more specifically the archival description facing these changes and proposals, in order to illustrate the challenges faced by Contemporary Archival Science in a new context of production, organization and representation of archival knowledge.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 1987) [ClassicSimilarity], result of:
      0.18536724 = score(doc=1987,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 1987, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=1987)
  0.25 = coord(1/4)

Content

Introduction: envisioning semantic information spacesIndexing and knowledge organization -- Semantic technologies for knowledge representation -- Information retrieval and knowledge exploration -- Approaches to handle heterogeneity -- Problems with establishing semantic interoperability -- Formalization in indexing languages -- Typification of semantic relations -- Inferences in retrieval processes -- Semantic interoperability and inferences -- Remaining research questions.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 2058) [ClassicSimilarity], result of:
      0.18536724 = score(doc=2058,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 2058, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=2058)
  0.25 = coord(1/4)

Content

Für die Software, vgl.: http://radimrehurek.com/gensim/index.html. Für eine Demo, vgl.: http://dml.cz/handle/10338.dmlcz/100785/SimilarArticles.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 2133) [ClassicSimilarity], result of:
      0.18536724 = score(doc=2133,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 2133, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=2133)
  0.25 = coord(1/4)

Abstract

A new method for visualizing the relatedness of scientific areas has been developed that is based on measuring the overlap of researchers between areas. It is found that closely related areas have a high propensity to share a larger number of common authors. A method for comparing areas of vastly different sizes and to handle name homonymy is constructed, allowing for the robust deployment of this method on real data sets. A statistical analysis of the probability distributions of the common author overlap that accounts for noise is carried out along with the production of network maps with weighted links proportional to the overlap strength. This is demonstrated on 2 case studies, complexity science and neutrino physics, where the level of relatedness of areas within each area is expected to vary greatly. It is found that the results returned by this method closely match the intuitive expectation that the broad, multidisciplinary area of complexity science possesses areas that are weakly related to each other, whereas the much narrower area of neutrino physics shows very strongly related areas.

0.04634181 = product of:
  0.18536724 = sum of:
    0.18536724 = weight(_text_:handle in 3263) [ClassicSimilarity], result of:
      0.18536724 = score(doc=3263,freq=2.0), product of:
        0.42740422 = queryWeight, product of:
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.06532823 = queryNorm
        0.43370473 = fieldWeight in 3263, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          6.5424123 = idf(docFreq=173, maxDocs=44421)
          0.046875 = fieldNorm(doc=3263)
  0.25 = coord(1/4)

Abstract

In this article, the use of a new term extraction method for query expansion (QE) in text retrieval is investigated. The new method expands the initial query with a structured representation made of weighted word pairs (WWP) extracted from a set of training documents (relevance feedback). Standard text retrieval systems can handle a WWP structure through custom Boolean weighted models. We experimented with both the explicit and pseudorelevance feedback schemas and compared the proposed term extraction method with others in the literature, such as KLD and RM3. Evaluations have been conducted on a number of test collections (Text REtrivel Conference [TREC]-6, -7, -8, -9, and -10). Results demonstrated that the QE method based on this new structure outperforms the baseline.